Data Scientist & ML Engineer
Abhijeeth Erra
I build modeling systems that improve real-world infrastructure.
I work at the intersection of data, systems, and impact. My background in neurobiology taught me to think in circuits and feedback loops; I now apply that lens to ML pipelines, evaluation frameworks, and infrastructure that has to work under real-world constraints. I care about models that are interpretable, robust, and built for production.
Featured Projects
Systems and analyses built for clarity and deployment.
ABRA
ML pipeline for automated analysis of Auditory Brainstem Responses (ABRs) in mice: preprocessing + Convolutional Neural Network (CNN) models for automated peak finding and threshold detection. Streamlit app for batch upload and metrics/visualizations export.
ETA Accuracy & Route Quality Analysis
End-to-end analytics for ETA accuracy using NYC Taxi trip data: PyTorch model, PostgreSQL, segmentation metrics, and failure-mode analysis.
On Modeling & Impact
Good modeling is not about fitting curves—it is about encoding assumptions explicitly, testing them, and designing systems that fail gracefully. I am drawn to problems where the stakes are high and the data is messy: healthcare, transportation, and public infrastructure. In those domains, a model is only as good as its integration with the rest of the stack and the clarity of its limitations.
I believe in building for the long term: documentation, evaluation frameworks, and tradeoff reasoning that outlast any single model version. The goal is not to impress with complexity but to deliver reliable, interpretable systems that teams can trust and iterate on.